The antileukemia activity of natural product HQ17(3) is possibly associated with downregulation of miR-17-92 cluster.

The compound 10'(Z),13'(E),15'(E)-heptadecatrienylhydroquinone [HQ17(3)] was purified from the sap of the lacquer tree Rhus succedanea. HQ17(3) has cytotoxic effect on cancer cells and can inhibit topoisomerase (topo) IIα activity. We treated various cancer cells with different doses of HQ17(3) and found that leukemia cells were most sensitive to HQ17(3). After analysis of microRNA (miRNA) profiling, we found that treatment with HQ17(3) caused downregulation of miR-17-92 cluster in some leukemia cells. These changes partially restored the normal levels from leukemia-specific miRNA expression signature. Messenger RNAs of tumor suppressor proteins, such as pRB, PTEN, and Dicer, are targets of miR-17-92 cluster. Their protein levels were increased after the treatment. c-Myc is a regulatory protein for miR-17-92 gene. Similar to topo IIα, we found that c-Myc decreased its activity after the HQ17(3) treatment, which may explain the downregulation of miR-17-92 cluster. Combined with 5-fluorouracil, NaAsO2, or ABT-737, HQ17(3) elicited additive inhibitory effects on leukemia cells. In conclusion, the high sensitivity of leukemia cells to HQ17(3) may be associated with the reduction of topo IIα and c-Myc activities, as well as with the downregulation of the miR-17-92 cluster expression.

Comparison between the repression potency of siRNA targeting the coding region and the 3'-untranslated region of mRNA.

Small interfering RNAs (siRNAs) are applied for post-transcriptional gene silencing by binding target mRNA. A target coding region is usually chosen, although the 3'-untranslated region (3'-UTR) can also be a target. This study elucidates whether the coding region or 3'-UTR elicits higher repression. pFLuc and pRLuc are two reporter plasmids. A segment of FLuc gene was PCR-amplified and inserted behind the stop codon of the RLuc gene of the pRLuc. Similarly, a segment of RLuc gene was inserted behind the stop codon of FLuc. Two siFLuc and two siRLuc were siRNAs designed to target the central portions of these segments. Therefore, the siRNA encountered the same targets and flanking sequences. Results showed that the two siFLuc elicited higher repression when the FLuc segment resided in the coding region. Conversely, the two siRLuc showed higher repression when the RLuc segment was in the 3'-UTR. These results indicate that both the coding region and the 3'-UTR can be more effective targets. The thermodynamic stability of the secondary structures was analyzed. The siRNA elicited higher repression in the coding region when the target configuration was stable, and needed to be solved by translation. A siRNA may otherwise favor the target at 3'-UTR.

Prunellae spica extract contains antagonists for human endothelin receptors.

Endothelin-1 (ET-1) is a powerful vasoconstrictor that contributes to blood pressure elevation. The biological effects of ETs are mediated by two receptors, namely, endothelin type A receptor (ET(A)R) and endothelin type B receptor (ET(B)R). Chinese herbal medicines (CHM) with antagonist activity for these two receptors were screened by establishing stable clones of CHO-K1 cells expressing high levels of human ET(A)R and ET(B)R, namely CHO-ET(A)R and CHO-ET(B)R.The aqueous extract of Prunellae Spica (P1) inhibited the binding of (125)I-ET-1 to ET(A)R and ET(B)R in CHO-ET(A)R and CHO-ET(B)R cells, respectively. P1 suppressed the ET-1-induced mobilization of intracellular Ca(2+) . Through the alcohol fractionation of P1, the antagonists of human ET(A)R and ET(B)R were found to belong to different, separable ingredients and the antagonist of ET(A)R is more soluble in alcohol. The two antagonists were also effective in the test on human primary cells, HASMC and HUVEC. P1 successfully prevented the development of ET-1-associated hypertension in rats without further purification. These results indicate the presence of anti-hypertensive ingredients in P. Spica extract, at least through the inactivation of ET(A)R and/or ET(B)R.

Coding region of adiponectin contains a silent intron reactivated by the adjacent intervening sequence of vector.

Precise splicing pre-mRNA into correct mRNA is a tightly orchestrated process involving both cis and trans factors. However, the regulatory mechanism underlying alternative splicing remain elusive. An alternative splicing was revealed by comparing RT-PCR products (cDNA) of human adiponectin gene (ADPN) genes and sequencing the corresponding cDNA recovered from CHO-K1 cells transfected with a pIRES-neo vector carrying the cDNA. We determined that an 88-nt sequence in the original cDNA was missing from the adiponectin mRNA isolated from the transfected cells. After analyzing the flanking sequences and context of the 88-nt fragment, we discovered that it does have a typical intron configuration containing the splicing donor and acceptor, polypyrimidine tract, and branch site. A point mutation at the acceptor site (AG→TG) abolishes this splicing site indicating that it is a bona fide intron. The intron splicing defaulted again when the adjacent intervening sequence (IVS) of pIRES-neo was deleted or adiponectin 3'-UTR was present. We found that 3'-UTR segment contained several splicing silencers and IVS contained high density of splicing enhancers. It explained the reactivation of this silent intron. Our results elicited the possibility that a 3'-UTR-free cDNA may reactivate an otherwise silent intron in the coding region as it is cloned for expression in mammalian cells.

A cysteine-reactive alkyl hydroquinone modifies topoisomerase IIα, enhances DNA breakage, and induces apoptosis in cancer cells.

We previously reported that the anticancer activity of a botanical compound 10'(Z),13'(E),15'(E)-heptadecatrienylhydroquinone [HQ17(3)] was attributed to topoisomerase (Topo) IIα poisoning and the induction of oxidative damage. HQ17(3) irreversibly inhibits Topo IIα activity in vitro and is more cytotoxic in leukemia HL-60 cells than in Topo IIα-deficient variant HL-60/MX2 cells, which suggests that Topo IIα is a cellular target of HQ17(3). This study further characterizes the molecular mechanisms of the anticancer activity of HQ17(3). Proteomic analyses indicated that HQ17(3) reacted with Cys-427, Cys-733, and Cys-997 of recombinant Topo IIα in vitro, whereas it reacted with Cys-427 of cellular Topo IIα in Huh7 hepatoma cells. The modification of HQ17(3) inhibited Topo IIα catalytic activity, increased the Topo IIα-DNA cleavage complex, and caused the accumulation of DNA breakage. In Huh7 cells, HQ17(3) treatment caused prompt inhibition of DNA synthesis and consequently induced the expression of DNA damage-related genes DDIT3, GADD45A, and GADD45G. Topo IIα inhibition, apoptosis, and oxidative stress were found to account for cytotoxicity caused by HQ17(3). Pretreatment of Huh7 cells with N-acetylcysteine (NAC) partially attenuated mitochondrial membrane damage, DNA breakage, and caspase activation. However, NAC pretreatment did not diminish HQ17(3)-induced cell death. These results suggest that the anticancer activity of HQ17(3) is attributed significantly to Topo IIα poisoning. The structural feature of HQ17(3) can be used as a model for the design of Topo IIα inhibitors and anticancer drugs.

Potential effect of resistin on the ET-1-increased reactions of blood pressure in rats and Ca2+ signaling in vascular smooth muscle cells.

Resistin and endothelin-1 (ET-1) are upregulated in people with type II diabetes mellitus, central obesity, and hypertension. ET-1 signaling is involved in Ca(2+)-contraction coupling and related to blood pressure regulation. The aim of this study is to investigate the role of resistin on ET-1-increased blood pressure and Ca(2+) signaling. The blood pressure and cytosolic Ca(2+) of vascular smooth muscle cells (VSMCs) of Sprague-Dawley rats were detected. The data demonstrated that resistin accelerated and prolonged ET-1-induced increases in blood pressure and had significant effects on ET-1-increased Ca(2+) reactions. Resistin-enhanced ET-1-increased Ca(2+) reactions were reversed by blockers of store-operated Ca(2+) entry (SOCE) and extracellular-signal-regulated kinase (ERK). The endogenous expression of Orai and stromal interaction molecular (STIM) were characterized in the VSMCs. Furthermore, resistin-enhanced ET-1 Ca(2+) reactions and the resistin-dependent activation of SOCE were abolished under STIM1-siRNA treatment, indicating that STIM1 plays an important role in resistin-enhanced ET-1 Ca(2+) reactions in VSMCs. Resistin appears to exert effects on ET-1-induced Ca(2+) increases by enhancing the activity of ERK-dependent SOCE (STIM1-partcipated), and may accelerate and prolong ET-1-increased blood pressure via the same pathway.

Systemic comparison of repression activity for miRNA and siRNA associated with different types of target sequences.

Most miRNA target sites are determined by Watson-Crick pairing via the seed region (positions ∼2-7nt of the mature miRNA). The binding sites of target mRNAs are categorized primarily as 7mer-m8, the 7mer-1A and the 8mer sites. This study analyzed post-transcriptional repression as a function of associations among various seed/target sequences. The target sequence of miR-155 from TP53INP1 was modified such that their various monomers and dimers could be inserted into the 3'UTR of a reporter gene for monitoring repression activity of miR-155. Results revealed that the level of repression could be ordered as follows: perfect-matched target≫dimeric targets≫monomeric targets. For dimeric targets, the order is 2×8mer>2×7mer-m8>2×7mer-1A. Fold repression of 8mer+7mer-1A lay between 2×8mer and 2×7mer-1A. A mismatch in one seed dramatically decreasing repressive activity of the dimer. This indicated that the degree of repression could be synergistically enhanced through the cooperation of the two miRISC-loaded monomers. The siRNA-155 (siRNA carrying miR-155 sequence) elicited higher repressive activity than miR-155, as they bound to the perfectly matched target. However, strong repression of miR-155 and siRNA-155 with a perfectly matched target was due primarily to translational attenuation. Cleavage/degradation of the target mRNA was not a major cause of the observed repression.

Single point mutation of microRNA may cause butterfly effect on alteration of global gene expression.

MicroRNAs (miRNAs) are short nucleotide RNAs that negatively regulate gene expression at the post-transcriptional level. Partially double-stranded miRNAs interact with an RNA-induced silencing complex (RISC) where one strand termed the guide strand is selected, while the partner strand accumulates at a lower level and is presumed to be degraded. The miRNA-loaded RISC then binds to target mRNAs through imperfect complementary sequences located in the 3'UTR and causes translation inhibition. One miRNA may negatively regulate hundreds of target mRNAs. In this study, a pre-miR-155 mutant was used to elucidate that a single mutation creating a mismatch near the 3' end of miR-155 led to a shift in strand selection, causing an increased selection of miR-155(∗) and a decreased selection of miR-155, thereby fine-tuning the translation of their target genes. Consequently, this resulted in a butterfly effect on global gene expression. Indeed, nearly half of the genes we analyzed in this study showed altered expression. Provided that over 800 miRNAs have been identified in humans to date, mutation of miRNA is expected to play a critical role in species evolution and individual diversity.

Insulin up-regulates heme oxygenase-1 expression in 3T3-L1 adipocytes via PI3-kinase- and PKC-dependent pathways and heme oxygenase-1-associated microRNA downregulation.

Heme oxygenase-1 (HO-1), the rate-limiting enzyme in heme catabolism, has antioxidant, antiinflammatory, and antiapoptotic effects in many physiological systems. HO-1 activity in obese mice is lower than in controls, and a sustained increase in HO-1 protein levels ameliorates insulin resistance and compensatory hyperinsulinemia. In the present study, we explored the regulatory effect of insulin on HO-1 expression in 3T3-L1 adipocytes and the underlying mechanism. We investigated the time- and dose-effect of insulin on HO-1 expression in 3T3-L1 adipocytes. Using specific inhibitors acting on insulin signaling pathways, we clarified the involvement of insulin downstream signaling molecules in insulin-regulated HO-1 expression. We also investigated the involvement of microRNAs (miRNAs) in insulin-regulated HO-1 expression using microarray and real-time RT-PCR assays. In an in vivo study, we performed insulin/glucose coinfusion in rats to increase circulating insulin levels for 8 h, then measured adipocyte HO-1 expression. Insulin caused a significant increase in HO-1 expression that was time- and dose-dependent, and this effect was blocked by inhibition of phosphatidylinositol 3 (PI3)-kinase activation using LY294002 (50 µM) or of protein kinase C activation using Ro-318220 (2 µM), but not by an Akt inhibitor, triciribine (10 µM). Furthermore, incubation of 3T3-L1 adipocytes with 100 nm insulin resulted in a significant decrease in levels of the miRNAs mir-155, mir-183, and mir-872, and this effect was also blocked by pretreatment with LY294002 or Ro-318220, but not triciribine. An in vivo study in rats showed that 8 h of a hyperinsulinemic euglycemic state resulted in a significant increase in adipocyte HO-1 expression. In conclusion, insulin increases HO-1 protein expression in 3T3-L1 adipocytes via PI3-kinase and protein kinase C-dependent pathways and miRNAs down-regulation.

Arachidin-1, a peanut stilbenoid, induces programmed cell death in human leukemia HL-60 cells.

The stilbenoids, arachidin-1 (Ara-1), arachidin-3, isopentadienylresveratrol, and resveratrol, have been isolated from germinating peanut kernels and characterized as antioxidant and anti-inflammatory agents. Resveratrol possesses anticancer activity, and studies have indicated that it induces programmed cell death (PCD) in human leukemia HL-60 cells. In this study, the anticancer activity of these stilbenoids was determined in HL-60 cells. Ara-1 had the highest efficacy in inducing PCD in HL-60 cells, with an approximately 4-fold lower EC50 than resveratrol. Ara-1 treatment caused mitochondrial membrane damage, activation of caspases, and nuclear translocation of apoptosis-inducing factor, resulting in chromosome degradation and cell death. Therefore, Ara-1 induces PCD in HL-60 cells through caspase-dependent and caspase-independent pathways. Ara-1 demonstrates its efficacy as an anticancer agent by inducing caspase-independent cell death, which is an alternative death pathway of cancer cells with mutations in key apoptotic genes. These findings indicate the merits of screening other peanut stilbenoids for anticancer activity.

Short-term regulation of tumor necrosis factor-alpha-induced lipolysis in 3T3-L1 adipocytes is mediated through the inducible nitric oxide synthase/nitric oxide-dependent pathway.

TNF-alpha has several effects on adipocytes that may be related to the development of type 2 diabetes in obese subjects. Many studies demonstrated that long-term treatment with TNF-alpha increases lipolysis in adipocytes. However, the short-term (<4 h) effects of TNF-alpha on lipolysis have not been well investigated. The aim of this study was to investigate the short-term regulatory mechanism of TNF-alpha-induced lipolysis in 3T3-L1 adipocytes. Well-differentiated 3T3-L1 adipocytes were used. Lipolysis was determined by measuring glycerol release. Expression of inducible nitric oxide (iNOS) and nitric oxide (NO) production were measured, respectively, by Western blots and the Griess reagent. A selective iNOS inhibitor (s-ethylisothiourea . HBr), an adenylyl cyclase inhibitor (SQ22536), and a guanylyl cyclase inhibitor (LY83583) were used to investigate the involvement of iNOS, cAMP, and cGMP in TNF-alpha-induced lipolysis. Transient transfection with iNOS short hairpin RNA was performed to confirm the involvement of iNOS in TNF-alpha-induced lipolysis. Phosphorylation of hormone-sensitive lipase (HSL) was measured by immunoprecipitation and Western blotting. Results showed that short-term TNF-alpha treatment significantly increased lipolysis, iNOS expression, and NO production in a time- and dose-dependent manner. Furthermore, treatment with the NO donor S-nitroso-N-acetylpenicillamine also stimulated lipolysis and HSL phosphorylation in 3T3-L1 adipocytes. Moreover, pretreatment with inhibitors of iNOS and guanylate cyclase, but not an adenylate cyclase inhibitor, abolished TNF-alpha-induced lipolysis and HSL phosphorylation. Suppression of TNF-alpha-induced iNOS expression using short hairpin RNA significantly reduced TNF-alpha-induced lipolysis. In conclusion, short-term TNF-alpha treatment induces lipolysis in 3T3-L1 adipocytes by increasing iNOS expression and NO production, which activates the guanylyl cyclase/cGMP-dependent pathway and induces phosphorylation of HSL.

A multisampling reporter system for monitoring microRNA activity in the same population of cells.

MicroRNAs (miRNAs) downregulate gene expression by binding to the partially complementary sites in the 3' untranslated region (UTR) of target mRNAs. Several methods, such as Northern blot analysis, quantitative real-time RT-PCR, microarray, and the luciferase reporter system, are commonly used to quantify the relative level or activity of miRNAs. The disadvantage of these methods is the requirement for cell lysis, which means that several sets of wells/dishes of cells must be prepared to monitor changes in miRNA activity in time-course studies. In this study, we developed a multisampling reporter system in which two secretable bioluminescence-generating enzymes are employed, one as a reporter and the other as an internal control. The reporters consist of a pair of vectors containing the Metridia luciferase gene, one with and one without a duplicated miRNA targeting sequence at their 3'UTR, while the other vector coding for the secreted alkaline phosphatase gene is used as an internal control. This method allows miRNA activity to be monitored within the same population of cells over time by withdrawing aliquots of the culture medium. The practicability and benefits of this system are addressed in this report.

Specific repression of mutant K-RAS by 10-23 DNAzyme: sensitizing cancer cell to anti-cancer therapies.

Point mutations of the Ras family are frequently found in human cancers at a prevalence rate of 30%. The most common mutation K-Ras(G12V), required for tumor proliferation, survival, and metastasis due to its constitutively active GTPase activity, has provided an ideal target for cancer therapy. 10-23 DNAzyme, an oligodeoxyribonucleotide-based ribonuclease consisting of a 15-nucleotide catalytical domain flanked by two target-specific complementary arms, has been shown to effectively cleave the target mRNA at purine-pyrimidine dinucleotide. Taking advantage of this specific property, 10-23 DNAzyme was designed to cleave mRNA of K-Ras(G12V)(GGU-->GUU) at the GU dinucleotide while left the wild-type (WT) K-Ras mRNA intact. The K-Ras(G12V)-specific 10-23 DNAzyme was able to reduce K-Ras(G12V) at both mRNA and protein levels in SW480 cell carrying homozygous K-Ras(G12V). No effect was observed on the WT K-Ras in HEK cells. Although K-Ras(G12V)-specific DNAzymes alone did not inhibit proliferation of SW480 or HEK cells, pre-treatment of this DNAzyme sensitized the K-Ras(G12V) mutant cells to anti-cancer agents such as doxorubicin and radiation. These results offer a potential of using allele-specific 10-23 DNAzyme in combination with other cancer therapies to achieve better effectiveness on cancer treatment.

The use of 10-23 DNAzyme to selectively destroy the allele of mRNA with a unique purine-pyrimidine dinucleotide.

10-23 DNAzyme is an oligodeoxyribonucleotide-based ribonuclease. It consists of a 15-nt catalytic domain flanked by two target-specific complementary arms. It has been shown to cleave target mRNA effectively at purine (R)-pyrimidine (Y) dinucleotide. Taking advantage of this specific property, 10-23 DNAzyme was designed to cleave mRNA of a given allele at a unique RY dinucleotide while leaving the mRNA encoded from other alleles of the same gene intact. In this study, a p53-R249S (AGG-->AGT) mutant was tested. 10-23 DNAzyme was used to cut mutant mRNA at GT dinucleotide of codon 249. Both in vitro and in vivo studies showed that this DNAzyme could specifically cut the mutant p53 allele, leaving the wild-type unaffected. This proof-of-concept experiment provided a new way to knock down expression of a given allele with special single-base transversion.

An endothelin type A receptor-expressing cell to characterize endothelin-1 binding and screen antagonist.

Endothelin-1 (ET-1) induces contraction of vascular smooth muscle through binding to endothelin type A receptor (ET(A)R). COS-7 cells stably expressing high levels of the ET(A)R were established (designated COS-7(ET(A)R)). The COS-7(ET(A)R) cell bound [(125)I]ET-1 with a K(d) of 932+/-161 pM and a B(max) of 74+/-13 fmol/2x10(5) cells. [(125)I]ET-1 binding was inhibited by ET-1 and the ET(A)R antagonist BQ-610, but not by the endothelin type B receptor (ET(B)R) antagonist BQ-788. In clones expressing two ET(A)R mutants containing D46N or R53Q substitutions in the first extracellular domain of the receptor, [(125)I]ET-1 binding activity was dramatically reduced. This suggests that these single amino acid substitutions alter the three-dimensional structure of the ligand-binding domain of the ET(A)R. Using COS-7(ET(A)R) cell, we showed that Ca(2+) or Mg(2+) was essential for ET-1 binding to the ET(A)R and that ET-1 treatment induced postreceptor signaling, that is, intracellular accumulation of cyclic AMP (cAMP) and Ca(2+) mobilization. The COS-7(ET(A)R) established in this study will be a useful tool for screening ET-1 antagonists for treating hypertension.

Mammalian cis-reporting plasmid may alter activities due to the derivation of host Escherichia coli strains.

Methylation status of CpG dinucleotides in the promoter/regulatory region contributes to regulation of transcriptional activities of downstream genes. Nearly all plasmid vectors used in mammalian cells are generated from transformed Escherichia coli. However, these E. coli hosts may have different DNA methylation activities. For instance E. coli JM109 and DH5alpha contain Dam and Dcm methylases, which are absent in E. coli JM110 and GM2163. It has not been determined whether plasmids propagated from E. coli of different methylation activities result in altered expression in mammalian cells when transient transfection is conducted. In this report, cis-reporting plasmids were tested. When promoter/enhancer of tested plasmids contained several Dam/Dcm sites, the cis-reporting activity was 2 to 3 fold lower for those plasmids isolated from JM109 than from JM110. In contrast, the E. coli-derived methylation had little effect on transcription when the sites of methylation resided in the coding region. These findings suggest that cis-reporting plasmids used in comparative or successive experiments are required to be derived from the E. coli strain of the same methylation status. The plasmid for promoter-transcription factor studies should be Dam/Dcm negative E. coli strain.

Facilitated in vivo synthesis of ribonucleic acid and protein via T7 RNA polymerase.

Ribozyme and small interfering RNA (siRNA) now are widely used to suppress target genes bearing homologous sequences. In this study, commonly used cell lines (e.g., HEK, HeLa, H1299) were stably transfected with gene encoding T7 RNA polymerase. The cytoplasm-restricted transcription activity of T7 RNA polymerase confers a continuous and robust transcription from T7 promoter-containing oligonucleotide (ODN) template for siRNA or ribozyme and leads to 70 to 80% inhibition of the tested target genes. ODN template offers the advantages of being more stable and economical than synthetic or in vitro-transcribed siRNA or ribozyme. Compared with the use of siRNA/ribozyme-expressing plasmids, our system does not require procedures with preparations of recombinant plasmids and enrichment of transfected cells and can be applied to synthesize protein in which different levels of translation could be modulated via variations in the presence of polyA tail or internal ribosome entry site (IRES) in the T7-transcribed RNAs. The results of our current study provide a rapid and efficient system for the assay of in vivo synthesis and expression of RNAs and proteins.

Methylation profiling using degenerated oligonucleotide primer-PCR specific for genome-wide amplification of bisulfite-modified DNA.

DNA methylation is one of the essential epigenetic processes that play a role in regulating gene expression. Aberrant methylation of CpG-rich promoter regions has been associated with many forms of human cancers. The current method for determining the methylation status relies mainly on bisulfite treatment of genomic DNA, followed by methylation-specific PCR (MSP). The difficulty in acquiring a methylation profiling often is limited by the amount of genomic DNA that can be recovered from a given sample, whereas complex procedures of bisulfite treatment further compromise the effective template for PCR analysis. To circumvent these obstacles, we developed degenerated oligonucleotide primer (DOP)-PCR to enable amplification of bisulfite-modified genomic DNA at a genome-wide scale. A DOP pair was specially designed as follows: first 3' DOP, CTCGAGCTGHHHHHAACTAC, where H is a mixture of base consisting of 50% A, 25% T, and 25% C; and second 5' DOP, CTCGAGCTGDDDDDGTTTAG, where D is a mixture of base consisting of 50% T, 25% G, and 25% A. Our results showed that bisulfite-modified DNAs from a cell line, cord blood cells, or cells obtained by laser capture microdissection were amplified by up to 1000-fold using this method. Subsequent MSP analysis using these amplified DNAs on nine randomly selected cancer-related genes revealed that the methylation status of these genes remained identical to that derived from the original unamplified template.

Methods for enrichment of a mRNA isoform with specific alternative splicing.

Alternative splicing of pre-mRNA is an important mechanism for regulating gene expression in higher eukaryotes. Recent genomewide analyses of alternative splicing indicate that 40-60% of human genes have alternative splice isoforms, although some variants exist only in relatively low abundance. It has been shown that proteins of different functions can be produced by a diverse array of mRNA derived from a single pre-mRNA. Inevitably cloning and expression of the corresponding mRNAs (cDNA) constitute an essential step toward elucidating the function of such protein isoforms. Here, we propose methods for enriching an mRNA isoform, which is carried out before the RNA preparation is subjected to reverse transcription polymerase chain reaction and cloning. While the negative selection method destroys unwanted mRNA variants, the positive selection enriches the desired variant. Focusing on the mRNA of rat ADAR2 (adenosine deaminase 2 that acts on RNA) as an example, we have achieved 16- and 19-fold enrichment of the given mRNA variant via the proposed negative and positive selection, respectively. Single use or combined uses of our method facilitates the isolation and cloning of a minor mRNA variant. Moreover, these methods can also be used to determine whether two alternative splicing events taking place in a pre-mRNA are linked.